Linux - KernelThis forum is for all discussion relating to the Linux kernel.
Notices
Welcome to LinuxQuestions.org, a friendly and active Linux Community.
You are currently viewing LQ as a guest. By joining our community you will have the ability to post topics, receive our newsletter, use the advanced search, subscribe to threads and access many other special features. Registration is quick, simple and absolutely free. Join our community today!
Note that registered members see fewer ads, and ContentLink is completely disabled once you log in.
If you have any problems with the registration process or your account login, please contact us. If you need to reset your password, click here.
Having a problem logging in? Please visit this page to clear all LQ-related cookies.
Get a virtual cloud desktop with the Linux distro that you want in less than five minutes with Shells! With over 10 pre-installed distros to choose from, the worry-free installation life is here! Whether you are a digital nomad or just looking for flexibility, Shells can put your Linux machine on the device that you want to use.
Exclusive for LQ members, get up to 45% off per month. Click here for more info.
I just compiled and installed the new kernel (2.6.22), but when try to boot into it, X dies telling me that it failed to load the nvidia module. If I try to do modprobe nvidia I get the same message. Here's an excerpt from the X log file:
Quote:
(II) Setting vga for screen 0.
(**) NVIDIA(0): Depth 24, (--) framebuffer bpp 32
(==) NVIDIA(0): RGB weight 888
(==) NVIDIA(0): Default visual is TrueColor
(==) NVIDIA(0): Using gamma correction (1.0, 1.0, 1.0)
(**) NVIDIA(0): Option "MetaModes" "DFP: 1680x1050 +0+0"
(**) NVIDIA(0): Option "AddARGBGLXVisuals" "True"
(**) NVIDIA(0): Enabling RENDER acceleration
(II) NVIDIA(0): Support for GLX with the Damage and Composite X extensions is
(II) NVIDIA(0): enabled.
(EE) NVIDIA(0): Failed to load the NVIDIA kernel module!
(EE) NVIDIA(0): *** Aborting ***
(II) UnloadModule: "nvidia"
(II) UnloadModule: "ramdac"
(II) UnloadModule: "fb"
(EE) Screen(s) found, but none have a usable configuration.
Fatal server error:
no screens found
And now my 2.6.20-16 kernel has the same problem. Fortunately I still had a 2.6.20-15 which works. What was weird was that when I went into the older one, under Ubuntu "Restricted Drivers", the nvidia drivers were disabled. I enabled and when I rebooted into the old kernel, I had nvidia, but the other kernels still dont work.
Im not too familiar with this. I think I understand how modules work, but is the nvidia driver specific to a certain kernel? Do I need to get a newer version or something? Thanks for any help.
Closed source modules are kernel specific ... since you compiled the new kernel instead of using aptitude, the module did not get upgraded automatically.
The nv module should operate your card.
It is not unusual for a kernel update to break proprietary drivers. For this reason, there is a regular request for a binary API in linux, which is always rejected.
nvdia are usually pretty good at staying current though.
Last edited by Simon Bridge; 09-14-2007 at 11:06 AM.
All you need to do is rebuild your nvidia module. The module is built off of a specific set of kernel headers and such; if you do anything with your kernel (even, I think, just re-compiling the one you have) you need to re-build your nvidia module for it to load right.
I've had the same problem in the past, and where I found the solution was in a tutorial by Ubuntu expert Alberto Milone, whose website is at http://albertomilone.com/index.html
Look up the links to Nvidia problem-solving there and I believe you will be on your way to success.
You mean, you read his guide: http://albertomilone.com/latest_nvidia_udsf_feisty.html
... where he says: "Every time you change or upgrade your kernel you will have to reinstall the Nvidia driver, no matter which method you used to install the driver."
???
The guide has two "method 1"'s ... but the author also provides "envy", which I have been hearing nice things about.
Envy is what I used, and it worked instantly and I had tears in my eyes because I'd been struggling for three or four hours to try to get that card to work (again).
I definitely vote for Envy.
But I'm done "upgrading" my kernel indefinitely because every time I do something else fails.
And the new kernel fixed the one thing I'd hoped it would, which was that every time I unplugged my laptop from the wall while it was running, my processor clocks down to save power, but linux I guess didn't "notice" and so my clock, music, and videos would just run slower. Anyway, problem solved, maybe it has something to do with the tickless timer thing, either way I'm happy.
I just compiled and installed the new kernel (2.6.22), but when try to boot into it, X dies telling me that it failed to load the nvidia module. If I try to do modprobe nvidia I get the same message. Here's an excerpt from the X log file:
And now my 2.6.20-16 kernel has the same problem. Fortunately I still had a 2.6.20-15 which works. What was weird was that when I went into the older one, under Ubuntu "Restricted Drivers", the nvidia drivers were disabled. I enabled and when I rebooted into the old kernel, I had nvidia, but the other kernels still dont work.
Im not too familiar with this. I think I understand how modules work, but is the nvidia driver specific to a certain kernel? Do I need to get a newer version or something? Thanks for any help.
The nvidia driver matches the kernel version with the modules for the driver. These modules are found in /lib/modules. In my case I have 2.6.18.8-0.5-default as my kernel version and
drwxr-xr-x 3 root root 72 2007-06-21 16:32 2.6.18.2-34-default
drwxr-xr-x 4 root root 504 2007-07-20 22:06 2.6.18.8-0.3-default
drwxr-xr-x 4 root root 560 2007-09-02 22:55 2.6.18.8-0.5-default
are the module directories even showing the previous version for Opensuse 10.2. The latest one matches my kernel version. If you recompile a new version of the kernel then the matching modules don't exist yet for the nvidia drivers until you rerun the nvidia.....run file again as root which will create a new module directory in /lib/modules to match the latest kernel version. Done it lots of times and works every time though a bit of a bother because X has to be shut down. In your case it didn't even start. Nvidia recompiles against the newer include files that come with the source of the newer kernel.
I had the same issue. I upgraded the feisty kernel from 2.6.16 to 2.6.22 because I hope it would solve an issue with an SATA LG CD/DVD player/burner (which it did ).
I have a nvidia graphics card and used the binary package that they (NVIDIA ) provide on their website. Just booted the new kernel in a terminal (by booting to run level 3). Installed the package again, rebooted and done.
Worked really well for me and it don't mind to redo it every time I install a new kernel. It's not like I'm installing new kernels every day.
You're right, ubuntu doesn't have a runlevel 3. What I meant is booting to a console without X. In ubuntu this is runlevel 1 (single user mode).
I'm sorry, probably too used to runlevel 3 in other distros. I don't understand anyway why Ubuntu has changed this. Anyone?
In all other distros I know, basically AFAIK it has always been like that: rl 3 = full user mode without X, rl5 = full user mode with X. I ignore which good reason Canonical has to change this standard behavior.
Anyway,
Ubuntu has only 2 runlevels (other than 0 and 6): 1 and 2.
You can boot Ubuntu into a console by doing the following:
1. Press ESC when system boots, to enable the grub menu
2. select the kernel that you want to boot using the arrow keys
3. Press 'e' to edit the command
4. Go down one line, to the line that starts with "kernel ...."
5. Press 'e' to edit
6. remove 'quiet' and 'splash' (optional) and add 'single' (no quotes, of course)
7. Press ENTER
8. Press 'b' to boot
For the second question, I know that Ubuntu has packages that provide nvidia packages, but:
1. they didn't work for me (the 'nv' driver in the xorg.conf that is), so I (am used to ) run of to the nvidia website to download the latest driver.
2. I had to install a custom kernel (2.6.22-6, IIRC) in order to get my SATA DVD burner to be recognised.
3. I had trouble with my firewire card, or more precisely the dv1394 kernel module, which magically fixed itself with the newer kernel.
(I needed my dv1394 kernel module to create a raw1394 device in order to be able to use Kino with my camcorder. Apparently there is some "disagreement" between the kernel guys and the video editing guys about the raw device being a security issue)
BTW, I compiled Kino myself as well, because I found version 1.1.1 much better than 0.9.2.
IIRC, whenever I tried it I got a "Your hardware doesn't require any restricted drivers" message and it all stops there.
Admitted, I didn't try to install the package "by hand" (via apt-get).
LinuxQuestions.org is looking for people interested in writing
Editorials, Articles, Reviews, and more. If you'd like to contribute
content, let us know.